Over the past few years, there has been a surge in real-world applications of AI with the rollout of technologies that use advances in learning, perception, and natural language. Today, there are vibrant ongoing discussions about how to maximize the value of these new applications and services, and about the potential societal influences of the technologies, including issues around ethics, economics, privacy, transparency, bias and inclusiveness, and trustworthiness of the technologies. The founding research scientists care about these issues too, and see the opportunity for ongoing cross-industry discussion and the development of best practices to help gain the most from AI technologies for the benefit of people and society.
This group foresees great societal benefits and opportunities ahead, but we also understand that as with every new technology there will be concerns and confusion associated with new applications and competencies. We look forward to working together on these important issues, including ethics, safety, transparency, privacy, biases, and fairness.
Amazon, Apple, DeepMind/Google, Facebook, IBM, and Microsoft have been developing AI-related technologies for years. Beyond their work on core R&D efforts, research scientists at these companies have been thinking about and discussing the potential societal impact of AI systems and how potential concerns might be addressed. Conversations have occurred at workshops, conferences, and smaller meetings. Earlier this year, several research scientists at the companies kicked off informal discussions about the possibility of bringing their companies together to form a non-profit organization charged with exploring and developing best practices. We are pleased those discussions culminated in the formation of the Partnership on AI and delighted to be adding six new Independent Trustees to our board.
The goals are to ensure that applications of AI are beneficial to people and society. We believe that artificial intelligence technologies hold great promise for raising the quality of people’s lives and can be used to help humanity address important global challenges. This organization seeks to ensure that our work fulfills these expectations. To do this, the organization will study the potential societal impact of AI systems, and develop and share best practices. We will also create working groups for different sectors, for example healthcare and transportation, allowing us to conduct research on the specific AI applications in these different sectors of the economy. We will also develop educational resources and host open forums to widely disseminate information about the latest topics in the field and support an ongoing public discussion about the technology.
We believe that by taking a multi-party stakeholder approach to identifying and addressing challenges and opportunities in an open and inclusive manner, we can have the greatest benefit and positive impact for the users of AI technologies. While the Partnership on AI was founded by six major IT companies, the organization will be overseen and directed by a diverse board that balances members from the founding companies with leaders in academia, policy, law, and representatives from the not-for-profit sector. By bringing together these different groups, we will also seek to bring open dialogue internationally, bringing parties from around the world to discuss these topics. A key operating principle is that we will share our work openly with the public and encourage their participation. The actions of the Partnership, including much of its discussions, meetings, results, and guidance, will be made publicly available.
A board of trustees will drive the activities of the Partnership. The board is comprised of six seats for representatives from each of the founding companies, and an equal number of seats for non-company representatives. An Executive Steering Committee, composed of working group leaders and others invited by the Board of Trustees, will guide, commission, and evaluate activities within the overall objectives and scope set by the Board. Day-to-day operations will be overseen by an executive director who will work closely with the board of directors. Conferences, meetings, panels, projects, and working groups will be commissioned by the board and conducted by the executive staff.
Each founding company will have a representative on the board. The current representatives are Ralf Herbrich from Amazon, Tom Gruber from Apple, Greg Corrado from DeepMind/Google, Yann LeCun from Facebook, Francesca Rossi from IBM, and Eric Horvitz from Microsoft. In addition to these founding members in January 2017 we welcomed six new Independent Trustees: Dario Amodei (OpenAI), Subbarao Kambhampati (AAAI/ASU), Deirdre Mulligan (UC Berkeley), Eric Sears (MacArthur Foundation), Carol Rose (ACLU) and Jason Furman (PIIE).
To meet its goals, the organization anticipates it will host discussions, commission studies, write and distribute reports on critical topics, and seek to develop and share best practices on the development, testing, and fielding of AI technologies. We will conduct outreach with the public and across the industry on topics related to advancing better understanding of AI systems and the potential applications and implications of this technology as they arise. Beyond addressing and engaging stakeholders on potential concerns, influences, and best practices, we will also work to identify new opportunities for employing AI advances for socially benevolent applications, and seek to identify areas of untapped opportunity for people and society.
We are also looking to appoint an Executive Director who will oversee the day-to-day operation of the organization. The Board of Trustees will meet as a whole for the first time in February 2017. We expect to be able to announce more details then, including how everyone can participate, as well as the initial program of research and activities.
We are very excited to work with anyone who is interested in joining our effort. This is a collaborative and multi-stakeholder organization and we want people with an interest in AI from across all fields to take part. If you want to take part in some way, please email us at email@example.com.
The Partnership on AI does not intend to lobby on behalf of the industry. It is designed to be an independent organization that passionately looks out for the short- and long-term futures of people and society more broadly. We foresee that guidance generated by shared, multi-party stakeholder discussions with balanced and independent reflection and analysis leading to best practices that, in turn, may be useful to industry and other sectors, including policymakers. Above all, our goal is that over time, our work will increasingly become a valued source of guidance, expertise, and authority on best practices for the ethical and safe development of AI.